Neoadjuvant Trastuzumab and Pertuzumab for Early HER2-Positive Breast Cancer: A Real World Experience

Background Randomized studies of neoadjuvant (NA) trastuzumab and pertuzumab combined with chemotherapy for HER2-positive breast cancers (BC) have reported pathological complete response (pCR) rates of 39 to 61%. This study aimed to determine the real-world efficacy and toxicity of NA trastuzumab and pertuzumab combined with chemotherapy in a UK tertiary referral cancer centre. Methods HER2-positive early BC patients given neoadjuvant chemotherapy with trastuzumab and pertuzumab between October 2016 and February 2018 at our tertiary referral cancer centre were identified via pharmacy records. Clinico-pathological information, treatment regimens, treatment-emergent toxicities, operative details, and pathological responses and outcomes were recorded. Results 78 female patients were identified; 2 had bilateral diseases and 48 of 78 (62%) were node positive at presentation. 55 of 80 (71%) tumours were ER-positive. PCR occurred in 37 of 78 (46.3%; 95% CI: 35.3–57.2%) patients. 14 of 23 (60.8%) patients with ER-negative tumours achieved pCR; 23 of 55 (41.8%) were ER-positive and 6 of 19 (31.6%) were ER-positive and PgR-positive. No cardiac toxicity was documented. Diarrhoea occurred in 53 of 72 (74%) patients. Grade 3–4 toxicity occurred in ≥2% patients. These were diarrhoea, fatigue, and infection. The Median follow up period was 45.2 months (95% CI 43.8–46.3) with 71 of 78 (91.0%) remaining disease-free and 72 of 78 (92.3%) alive. Estimated OS at 2 years 86% (95% CI: 75–99%). Conclusion This data confirms the efficacy of neoadjuvant chemotherapy combined with dual HER2 directed therapy. While no cardiac toxicity was observed, diarrhoea occurred frequently. The low pCR rate observed in ER and PgR-positive BCs warrants further investigation and consideration of strategies to increase the pCR rate.


Introduction
In HER2-positive breast cancer neoadjuvant studies have demonstrated that the addition of trastuzumab to chemotherapy improves pathological complete response rates compared to chemotherapy alone [1]. Subsequently, trials employing a number of treatment strategies have investigated the efficacy and tolerability of dual HER2 therapy added to chemotherapy in the neoadjuvant setting, including (1) trastuzumab in combination with chemotherapy versus trastuzumab plus pertuzumab with chemotherapy (NEOSPHERE and PEONY) [2,3]; (2) different chemotherapy regimens in combination with trastuzumab plus pertuzumab (BERENICE and TRYPHAENA) [4,5], and (3) trastuzumab plus pertuzumab with chemotherapy versus trastuzumab-emtansine (T-DM1) (KRISTINE) [6].
Primary endpoints varied, with NEOSPHERE, KRIS-TINE, and PEONY investigating pathological complete response (pCR) rate, while TRYPHAENA and BERENICE assessed the cardiac safety of HER2 blockade (Supplementary Table 1). e definition of pCR as a primary endpoint varied, with NEOSPHERE and PEONY defining it as a complete response in the breast and axilla (ypT0/is/N0), while in NEOSPHERE it was complete response in the breast alone (ypT0/is). e addition of pertuzumab to trastuzumab and chemotherapy in NEOSPHERE and PEONY observed a pCR (ypT0is/N0) rate of 39.3% compared to the control arms of trastuzumab and chemotherapy alone where rates of 21.5% and 21.8% were achieved [2,3] (Supplementary  Table 1). Across all trials, a lower pCR rate was noted for ERpositive tumours .3%) compared to ER-negative tumours (63. .4%) with the use of the doublet [2][3][4][5][6].
Given the known differences between the clinical trial environment and routine clinic setting, we sought to explore the efficacy and toxicity of neoadjuvant HER2 doublet in combination with chemotherapy within the practice of a tertiary referral NHS cancer centre in the UK.

Data Collection and Analysis.
e clinicopathological data collected included patient age, tumour information at time of diagnosis including radiological size, histological grade on core biopsy, histological subtype, axillary nodal status on biopsy, and estrogen receptor status. Assessment of progesterone receptor status was not performed in all cases as this is not considered mandatory, particularly in those of ER-status [12]. In the presence of multifocal disease within the same breast, total tumour size (T stage) was determined by the diameter of the largest lesion according to radiographic findings. In patients found to have bilateral disease each tumour was treated separately. Posttreatment information collected included date and type of surgery, histopathological information including size, grade, and axillary lymph node status. Echocardiogram reports were reviewed to assess LVEF. Details of the neoadjuvant chemotherapy regimen and temporal relationship to the initiation of trastuzumab-pertuzumab were collected.

HER2 Positivity.
All core biopsy reports from referring units were reviewed to confirm HER2 overexpression, as confirmed in a histopathological report by immunohistochemistry (IHC) and/or fluorescence in situ hybridization (FISH).

Treatment
Regimens. Treatment regimens consisted of (1) FEC-THP: three cycles of FEC100 (fluorouracil, epirubicin, cyclophosphamide) every 21 days followed by four cycles of docetaxel, trastuzumab, and pertuzumab. Where there were allergic reactions to docetaxel or there were tolerability issues, patients were switched to paclitaxel weekly.
Granulocyte colony stimulating factor (Filgrastim), postoperative radiotherapy and/or endocrine therapy as well as adjuvant zoledronic acid were prescribed as per local guidelines.

Study End-Point Definitions.
Pathological complete response was defined as the complete absence of invasive and in situ disease in the breast and the axillary node (ypT0/ is N0). For non-pCRs, residual tumour size was based on maximal diameter as described in the pathological report.
An event was defined as: ipsilateral invasive breast tumour recurrence, local/regional invasive breast cancer 2 e Breast Journal recurrence, distant recurrence, or death from any cause, whichever occurs first. Progression-free survival (PFS) was measured from date of the diagnostic core biopsy to first event, while disease-free survival (DFS) was measured from the date of definitive surgery to the first event. Patients who had not experienced an event were censored at the date of last follow-up when they were known to be event-free.
Overall survival (OS), measured from the date of the diagnostic core biopsy to death due to any cause. Survivors were censored at the date they were last known to be alive. e data cut-off was 6 th March 2021.
2.6. Adverse Events. Safety outcome measures included type, incidence, and severity of adverse events according to the Common Terminology Criteria for Adverse Events (CTCAE) criteria v1.1, documented by nursing staff at each treatment delivery visit. Cardiotoxicity was defined as a decrease in LVEF of 10% from baseline or a decrease to an LVEF of less than 55% at any time, as well as any Class 3 or 4 congestive heart failure according to the New York Heart Association [13].

Statistical Analysis.
Continuous variables are presented with their median and interquartile range (IQR), while categorical variables are described as frequency counts and proportion percentages. Fisher exact or Wilcoxon-Mann-Whitney tests were performed as appropriate to compare proportions and distributions. PCR was also shown graphically with bar plots and related 95% confidence intervals.

Results
Between October 2016 and February 2018, 79 patients received at least one dose of neoadjuvant trastuzumab and pertuzumab; clinico-pathological information is summarized in Table 1. One patient was found to have de novo metastatic disease and was therefore excluded, leaving 78 patients. All patients were female with a median age of 50 years (IQR: 44.4-60.2) and a median BMI of 27 kg/m 2 , (IQR: 23.2-29.4). 13 (17%) of 78 patients were less than 40 years of age. Most cases, 67 of 78 (86%), presented symptomatically. e median tumour size at presentation was 30 mm (IQR: 23.0 to 47.5) with 19 of 78 (24.4%) having multifocal disease. 4 of 78 (4.9%) were inflammatory BC. Two patients had bilateral disease, thus a total of 80 tumours were assessable for pathological assessment. 55 of 80 (68.8%) tumours were ER positive; PgR staining was performed in 33 of 55 (60%). 19 of 33 (58%) were PgR positive and 14 of 33 (42%) were PgR negative. 23 of 80 (28.8%) were confirmed ER-negative and PgR-negative. In patients with bilateral tumours, 1 had bilateral HER2 positive, ER positive and PgR-positive tumours while the other had 1 HER2 positive, ER positive and PgR positive and the other triple negative. 61 of 80 (76.3%) assessable axillae were biopsied at diagnosis. Of these, 5 of 61 (8%) had sentinel lymph node biopsy prior to commencing systemic therapy. 48 of 61 (86%) revealed axillary node involvement, with 1 patient having bilateral axillary node involvement, giving a total of 49 of 80 (61%) axillae involved. erefore, 49 of 78 (65.4%) patients had pathologically confirmed lymph node involvement at presentation.  e median time from diagnosis to treatment initiation was 4.9 weeks (IQR: 3.9-6.1). 11 of 78 (14%) patients did not receive all doses of HER2 therapy due to toxicity, proceeding to surgery. No patients discontinued treatment due to disease progression. Supplementary Figure 2 summarizes regimen alterations, dose reductions, and deferrals.
3.5. Toxicity. All patients experienced toxicity, with 11 of 78 (14%) experiencing grade 3 or higher. e most frequently occurring toxicity was fatigue, occurring in 71 of 72 (99%) patients. e most frequently occurring grade 3 or higher toxicity was infection, occurring in 4 of 72 (5.5%) patients, leading to 3 (75%) of these patients ceasing treatment early. Diarrhoea was documented in 53 of 72 (73.5%) patients, occurring at grade 3 or higher in 2 of 72 (3%) patients. e most common treatment-related toxicities are summarized in Table 3.

Discussion
In the neoadjuvant setting, chemotherapy in combination with trastuzumab and pertuzumab in HER2 positive early BC patients is associated with an increase in the pCR rate compared to trastuzumab alone, with the doublet resulting in rates of 39.3 to 60.7% [2][3][4][5][6]. It is recognized that the realworld setting differs from that of randomized trials and that there is a need to develop real-world data regarding novel agents [14]. e cohort identified was prospectively treated between October 2016 and June 2018, reflecting the initial period when neoadjuvant trastuzumab and pertuzumab were made available to patients in England via the CDF. Our population of early HER2-positive BC patients was broadly consistent with those in pivotal trials. We report a median age of 50 years; 68.8% of tumours being hormone receptor-positive and 86% LN positive, as compared to a median age of 49-50 years, hormone receptor-positive 48-62% and LN positive 70-76% in pivotal studies [2][3][4][5][6].
pCR (ypT0/is/N0) occurred in 46% (37 of 80) of tumours in our cohort; within the range of 39. 3  We report a higher pCR rate for ER-negative as compared to ER-positive BCs, in keeping with the published randomized studies which ranged from 63.2-79.4% in ERnegative compared to 26-57.3% in ER-positive. [2][3][4][5][6]   e Breast Journal those that were HR positive [15][16][17][18][19][20][21]. ER status was reported as a significant predictor of pCR on multivariate analysis in some studies [17,19]. We investigated the efficacy of trastuzumab and pertuzumab in relation to the ER and PgR status of tumours and found a pCR rate of 64.3% (95%CI: 39-89%) in ER-positive and PgR-negative tumours compared to 31.6% (95%CI: 11-52.5%) for ER-positive and PgR-positive tumours. Across all five pivotal trials, "positive HR status" is defined as either ER or PgR positivity, with no data reported for ERpositive/PgR-positive and ER-positive/PgR-negative tumours. Across RWS, two reported PgR status of tumours [18,19]; one of these found PgR status was a significant predictor of pCR on univariate analysis (p < 0.01) but not on multivariate analysis [19]. e absolute pCR rate according to PgR status was not reported. e lower pCR rate reported with chemotherapy and HER2 doublet in ER-positive/PgR-positive BCs raises questions regarding the optimal treatment backbone for such BCs. In the NA-PHER2 study, 30 ER-positive and HER2-positive breast cancers were treated with 18 weeks of trastuzumab and pertuzumab, combined with fulvestrant and palbociclib and were assessed for pCR (ypT0/is/N0) [22]. NA-PHER2 reported a pCR rate of 27% (95% CI: 12-46%) [22]. In the PerELISA study, a pCR rate of 20.5% (95%CI: 11.1-34.5%) following 15 weeks of trastuzumab, pertuzumab, and letrozole was reported [23]. None of these studies reported any data based on the PgR status. ese data demonstrate that a noncytotoxic approach combining HER2-directed and endocrine therapy with or without a CDK4/6 inhibitor might be an effective and less toxic treatment option for ER-positive and PgR-positive BCs in the neoadjuvant setting.
Since March 2019, within England, adjuvant trastuzumab and pertuzumab have been for patients with node-positive disease or axillary node scarring consistent with a response to treatment. While we did not examine for axillary node scarring, 21 of 41 (51.2%) patients that did not achieve pCR had residual nodal disease and would therefore have been eligible to have continued treatment. Similarly, adjuvant TDM-1 which has demonstrated a 50% lower risk of recurrence in those with residual disease at surgery [26] has also subsequently become available for use within England on the NHS, and 41 of 78 (63%) of patients in the current study where pCR was not achieved would have been eligible for T-DM1.
No cardiotoxicity was identified in our population as defined by either LVEF decline of ≥10% or symptom onset. ese data are reassuring as to the safety of HER2 therapy in the context of a real population. e lack of cardiotoxicity may reflect the fact that clinicians have been using HER2 directed therapy for some time and so are better equipped at selecting patients as well as optimizing cardiac function in those with pre-existing conditions in conjunction with cardio-oncology colleagues [27].
Diarrhoea of any grade was very common in our cohort, occurring in 73.6% of patients (Table 3), with only 2.6% experiencing grade 3 or higher, as compared to rates of 10.1%-73.5% and 0-15% for all grades and grade 3 or higher, respectively, from previous studies (Supplementary Table 4) [2][3][4][5][6]. While 14.1% patients did not receive all doses of HER2 directed therapy, in none of these instances was diarrhoea the cause of early treatment cessation. ese data also highlight the challenges of using docetaxel in the real world as demonstrated by the 27% (21 of 78) of patients who were switched to weekly paclitaxel or nab-paclitaxel.

Limitations.
is study has several limitations. Firstly, this data is retrospective. Although patients were identified from a prospective database, the population size is relatively small and is therefore inadequately powered for robust analysis between subgroups such as ER and PgR status. Furthermore, since this cohort began treatment, several alternative therapeutic options have become available for use in early HER2 BC including postoperative pertuzumab and adjuvant TDM-1 [26,28]. erefore, the outcome data reported here may differ from those of patients diagnosed

Conclusion
In conclusion, in the context of a tertiary referral NHS Cancer Centre neoadjuvant chemotherapy in combination with trastuzumab and pertuzumab is a safe and efficacious treatment for early HER2 breast cancer. ese data are in keeping with data from randomized controlled studies and similar RWS, however the lower pCR rate in ER-positive, PgR-positive tumours identified here warrants further investigation.

Data Availability
Data in this article is not stored in a publicly archived dataset due to the confidential nature.

Conflicts of Interest
CP received grant from Pfizer and Daiichi Sankyo and honoraria from Pfizer, Roche, Daiichi Sankyo, Novartis, Exact sciences, Gilead, SeaGen, and Eli Lilly.

Supplementary Materials
Supplementary  Table 2: characteristics of residual tumour in those patients who did not achieve pathological complete response. Samples insufficient for assessment were either too small (n � 3), of inadequate quality (n � 2) or were found to only have intravascular tumour emboli remaining in the breast (n � 1). Supplementary  Table 3: summary of patients who developed recurrent disease and the treatments they received; * All the patients have died; * * found to be triple negative on biopsy of recurrence; AC: adriamycin, cyclophosphamide; RT; radiotherapy. Supplementary Table 4: adverse events reported by published neoadjuvant and dual HER2 blockade trials. Rates reported include adverse events of all severity. Rates reported in this paper are documented in the far-right column. * 5.2% of patients in this cohort presented with either a Grade 3 or 4 infection during their course of treatment. Supplementary Table 5: summary of real-world studies examining the efficacy of neoadjuvant chemotherapy and pertuzumab/trastuzumab doublet. * Only overall pCR rates were reported in some papers but were not stratified according to both hormone receptor status and treatment regimen. NAC: neoadjuvant chemotherapy (taxanes with or without anthracyclines; regimen not specified). A: doxorubicin C: carboplatin; H: trastuzumab; P: pertuzumab; T: docetaxel Cy: cyclophosphamide Pa: paclitaxel NR: results were not specifically reported. (Supplementary Materials)